Semen is among the most frequently encountered biological materials in forensic casework, particularly in sexual assault investigations. While DNA profiling of sperm cells enables donor identification when reference profiles are available, additional investigative intelligence is required in cases lacking database matches. Age estimation based on sperm-specific DNA methylation patterns has emerged as a promising complementary approach.
ForSpEC (Forensic Sperm Epigenetic Clock) is a compact model developed for chronological age prediction based on DNA methylation at seven highly informative CpG sites. The model was trained using Illumina EPIC array data, achieving a mean absolute error (MAE) of 3.13 years and a high correlation between predicted and chronological age (R² = 0.7). Cross-platform validation demonstrated that ForSpEC retains high predictive accuracy when applied to targeted sequencing approaches, including Ion AmpliSeq technology, with only a slight decrease in performance (MAE = 3.24 vs. 3.76 years).

In the next step, a comprehensive forensic validation framework was implemented, including: (i) comparison of age prediction accuracy between purified sperm cells and whole semen samples, (ii) evaluation across different population groups, (iii) assessment of repeatability and accuracy, and (iv) benchmarking against alternative sperm-based epigenetic age predictors. The results demonstrated a significant impact of cellular composition on prediction accuracy, attributed to the sensitivity of marker cg14728380 to somatic cell-derived DNA (MAE = 10.22 vs. 3.08 years). To address this limitation, an alternative analytical approach was proposed, enabling accurate age prediction in samples affected by somatic cell contamination and improving model robustness. Benchmarking showed that ForSpEC achieved the second-best performance after the GLA model, despite relying on a substantially smaller number of CpG markers.

In addition, the impact of potential confounding factors was evaluated, including lifestyle variables (smoking, alcohol consumption, and BMI) and semen quality parameters (sperm concentration, motility, morphology). Overall, these factors had no or only a minor effect on methylation values. Small but statistically significant associations were observed for BMI and smoking at cg00534655 (p-value = 0.046, β = 0.095) and cg23335134 (p-value = 0.016, β = −0.115), respectively. However, these effects did not translate into changes in prediction accuracy.

Overall, the expanded validation confirms that ForSpEC is a robust and accurate tool for forensic age estimation from sperm cell DNA, with strong potential for operational implementation as an investigative intelligence approach.

Authors

• Aleksandra Pisarek-Pacek (Institute of Zoology and Biomedical Research, Jagiellonian University, Poland)
• Kamila Marszałek (Department of Genomics and Forensic Genetics, Pomeranian Medical University in Szczecin, Poland)
• Kamila Stanisz (Department of Genomics and Forensic Genetics, Pomeranian Medical University in Szczecin, Poland)
• Maria Stępień (Department of Genomics and Forensic Genetics, Pomeranian Medical University in Szczecin, Poland)
• Antonia Heidegger (Institute of Legal Medicine, Medical University of Innsbruck, Austria)
• Aneta Macur (PARENS Fertility Centre, Poland)
• Jarosław Janeczko (PARENS Fertility Centre, Poland)
• Gabriela Kania (The Fertility Partnership Macierzynstwo, Poland)
• Marta Sikora-Polaczek (The Fertility Partnership Macierzynstwo, Poland)
• Miguel Boullón-Cassau (Forensic Genetics Unit, Institute of Forensic Sciences, University of Santiago de Compostela, Spain)
• Mafalda Silva (Forensic Genetics Unit, Institute of Forensic Sciences, University of Santiago de Compostela, Spain)
• Adrián Ambroa-Conde (Forensic Genetics Unit, Institute of Forensic Sciences, University of Santiago de Compostela, Spain)
• Andrzej Ossowski (Department of Genomics and Forensic Genetics, Pomeranian Medical University in Szczecin, Poland)
• Walther Parson (Institute of Legal Medicine, Medical University of Innsbruck, Austria)
• Ana Freire-Aradas (Forensic Genetics Unit, Institute of Forensic Sciences, University of Santiago de Compostela, Spain)
• María Victoria Lareu (Forensic Genetics Unit, Institute of Forensic Sciences, University of Santiago de Compostela, Spain)
• Ewelina Pośpiech (Department of Genomics and Forensic Genetics, Pomeranian Medical University in Szczecin, Poland)
• Wojciech Branicki (Institute of Zoology and Biomedical Research, Jagiellonian University, Poland)